Planning an aspheric scratch build

luminarium iaculator, I just implied there was “a buttload of useless talk up there”, which is different than presuming “all of the above up there is a buttload of useless talk”. ;-)

The collimator just above the emitter could be a extremely strong custom made fresnel lens, big and thick.

Cheers

Have you ever tried using a lens to focus a COB LED?
You would know that the size of the emitter ALWAYS matters, and as long as it is not a point source of light, all those wikipedia drawings cannot be applied to real life.

As mentioned earlier, having a large thickness lens close to the LED will collect a lot of light yes but also make a huge spot.
Buy yourself a short focal distance lens and test it out yourself, it sucks for throw.

This depends on how you define throw. As long as the lens diameter is the same, the luminous intensity (candela) of the beam will be the same for different focal length lenses. I personally think having a larger spot is usually better. Isn’t putting as many lumens as possible downrange what this is all about?

Like calculating parallel resistance. (1/Rt) = (1/R1) + (1/R2)

True, but at the same time the point of having a concentrated point of light with no spill will kinda be defeated.
I prefer near-collimated light, so that the divergence is as close to 0 as possible.
It is also more complicated to make thick aspheres with a close focal length than thin aspheres with a long focal length.
The longer the focal length is, the more similar an aspheric lens gets to a spherical one.

It’s basically a compromise / personal preference between larger spot, more divergence, more lumens, vs smaller spot, less divergence, less lumens.

Later this week I am going to try the same math we did in the BLF GT thread to make a few graphs, there may or may not be an optimal point of lumens and divergence, since we determined that 45 degrees is where the most lumens are coming out of an LED.

Fair enough.

I like to use a precollimator lens in my aspheric throwers. It can easily double or triple the light collection and beam area (relative to a typical single lens) while reducing the throw by less than 10. Light collection efficiency into the beam can be around 70 with a simple precollimator setup.

It is not clear to me the advantage of using more than 2 lenses. I don’t think it can increase the throw as this is determined by the area of the front lens. So is it just to increase the efficiency even more? I would think the losses from additional lenses would tend to counteract this.

:+1:

Maybe sedstar can enlighten us with some of this complex “math” he speaks of, because so far there is no existing proof that a multi-lens setup can get more throw than a single lens of the same max diameter.

okay… okay… okay! (I am skimming the posts that appeared while i had my 8 or so hour much-needed nap)

Now? I am happy. Now we are cooking with gas!. Now, we have several people attracted, involved and engaged… all of whom are used to looking at basic equations.

a couple basic points i would like to make back?

—- Mr enderman can say whatever he wants to me, it doesnt bother me in ANY way. No ones “feelings” matter the slightest bit as far as to whether i am right or not right.
—- i like the how many potatoes joke back. i’m irish, and i have a sense of humor. good show!
—- what i think is confidence, anyone else could easily take for rude or conceited or arrogant. “semantics”
—- i write clear and concise enough english that all my information is completely conveyed. You keep insisting that perfect punctuation and grammar are what wins mathematical arguments? Let me know how that goes for you.

aha, here we go. Here’s a little gem:

“Maybe sedstar can enlighten us with some of this complex “math” he speaks of, because so far there is no existing proof that a multi-lens setup can get more throw than a single lens of the same max diameter.”

The equation isnt a particularly hard one? Its 8th grade algebra to solve it. It implies and states, among other things? that ANY lens with positive focal length? can be replaced by TWO OTHER LENSES with an airgap, such that the OVERALL focal length is exactly the same.

Necessarily, the rear lens will be CLOSER to the LED… it HAS to collect more light that wasnt lost by the original one lens being so far away. (you keep getting hung up on the idea that the SAME ORIGINAL LENS will then have less focal length? Perfectly true statement. If your original lens was 100mm FL? wonderful… simply choose 2 lenses of 150mm FL, as an example.

placed touching (no airgap)? they form a single lens of 75mm FL (which would seem to support enderman’s “always less” statements) BUT when you introduce the airgap? the focal length begins to RISE UPWARDS and approaches 150mm if you go far enough.

if you STOP increasing the airgap when the system hits 100mm FL? What have you achieved?

===

Build number 1? a single positive lens of 100mm FL is used. We ALL know the problem, but i state it anyways. The problem is that at that 100mm distance? a lot of light gets lost and doesnt hit the lens. Since you can only increase diameter to a certain practical extent? Youy are stuck like chuck with a “dance” between higher focal length, and the trade-off loss of lost light from the distance.

Build number 2? TWO lenses of 150mm FL e-a-c-h replace the single lens. At a certain airgap? They “hit” 100mm FL, which was the value of your original lens. The REAR lens is much closer to the LED, its losing less light that was lost. The overall FL is STILL 100mm FL in sum total.

no where did i state that any particular difficult to source lens was needed. ANY of us goes and buys a single 100mm FL lens, dont we? ANY of us can buy instead two 150mm FL lenses to replace the one lens, and have the same FL as we wanted originally, except now with one lens farther back, collecting and not wasting light.

choose the 2 lenses carefully? run the numbers enough? you can put things where you want them, and its always an advantage.

==

you want to SEE the math? right out in the open for you? hiding right out in PLAIN SIGHT ?

https://en.wikipedia.org/wiki/Lens\_(optics)

go down the page, down the page… right underneath “chromatic aberrations”? right below that is the “compound lens”. There are the 4 equations. If you sit and play with paper/pencil/calculator long enough? using the example i just used of my 2 lens build replacing the other 1 lens build? you will see it too… you just have to run it with the right airgap to “hit” it…

no ray traces,any 8th grade algebra student can handle it… and those equations? are hundreds of years old i bet,they are LAWS. they do not LIE. Perhaps instructive to note that there is no delta problem here, and CAMERA lenses seem to work in spite of the fact that Mr. Enderman seems hung up on things on his “ray tracing program” which I am glad i dont know about, they would prevent me from using paper and pencil to illustrate what i say.

LOOK, if you are going to use a 60mm diameter, 100mm long TUBE anyways? To have a 60mm diameter lens of 100mm FL anyways? You COULD use two 60mm diameter lenses, each with 150mm FL… and simply put THOSE into the tube. (you have an empty space of 60mm diameter, whats wrong with a second lens back there?)

THEN what Mr enderman seems hung up on, his “always shorter FL” hang-up? works right how you WANT IT to help you, instead of hurt you.

get off the idea of using the SAME lens with some “other” collimating lens behind it… get ONTO the idea of choosing two OTHER lenses that combine with airgap to REPLACE your original single lens.

the math? does not LIE. and? i’m npot the only one saying this.

Dr. JONES kept repeating himself saying it, out of hundreds of people arguing about “throw and beamshape” (meaningless drivel) that arent doing anything constructive? i was the only person that read that thread, and went and found what he was talking about.

THIS is how simple compound camera lenses work, they dont suffer from Mr Endermans “ray tracing perplexing delta issue”, do they? NO, they take pictures just fine.

THIS is why the “sunnranger” lens kicks butt, and has 4 lenses. THIS is why it has a higher FL than suggested by its short length.

THIS is what Dr Jones on CPF showed a picture of in the long trainwreck thread on his bench.

THIS is what the violin teacher thread someone posted the build is made from.

THIS is what they describe as “step two” in EDMUND OPTICS math white paper on “prototyping illumination systems”. step 1, is a single lens… step 2? two lenses, and the advantages!

https://www.edmundoptics.com/globalassets/resources/articles/prototyping-illumination-systems-with-stock-optical-components-en.pdf

=

its not just ME? all the experts are saying it.

people build this way. The engineers at edmunds optics, curiously dont describe the ray tracing delta paradox… i say? throw your ray tracing program away, its tell you something you dont need…

Look. almost everyone is doing “step 1” in the edmunds optics paper. Everyone is doing the same things, and expecting different results. insanity!

they describe the “throw problem, mannnnn” they just state it as MATH instead of undefined meaningless “rules” that dont seem to exist to me.

yeah, theres a limiting problem… edmunds optics? the very next thing, “step 2” immediately describes what i am saying to do… use 2 lenses to replace your single lens.

this is why i ended up with 2 IR illuminators that “throw more, man” than everyone ELSES build, and with slightly less diameter. One is 3 lenses the other is 4 lenses.

the advantages? are intuitive, and obvious.

(punctuation and grammar? do not win math arguments)

Say I have a 100mm diameter, 150mm focal length lens. (150mm away form the LED)

What two lenses <=100mm would you use (what is their focal length) and at what distances from the LED would you put them?

well, right before i take a STAB at that? Now that i at least have your attention, let me use that attention to “plop” the rest of the answer out.

you will notice that the “simple” equation to calculate the new effective focal length of the new compound lens to replace the original single lens? its all stated as mainly “reciprocals” which are a pain in the tookus to work with. you CAN “play” with different values of lenses to see what results come out of the playtime, but, its tedious. Also you have to plug in the BackFocalLength each and every time.

heres the same equation? stated the way you WANT to see it, and its much easier to plug in values and use them. I simply re-solved the equation again and again, until i didnt make any stupid algebra errors. The solution with all variables? It QUICKLY grows more and more complex, getting nuts… until right at the end? All the huge craziness starts to “cancel out” and it simplifies down…

new focal length = (F1 * F2) / (F1 + F2 - Airgap)

all units have to be identical (i suggest millimeters, for obvious reasons)

F1 = focal length of lens 1
F2 = focal length of lens 2
Airgap = distance of air between both lenses

you “plug in” to the BFL equation with each solution to see the new back focal length.

it should be obvious, that if the lenses are touching? the airgap is zero millimeters… and you are left with just the “products” divided by the “sums”. If you start using two DIFFERENT focal lengths? and start varying them each time you play? is when it really starts to get interesting…

==

Now that we are all c-a-l-m ? And you are at least willing to entertain my idea? I can now state… that i am sorry i had to act a little rude, and a little arrogant, to get your attention. I heartily apologize. But, you are the FIRST person, on three different sites, and of all the hundreds of people? That at least seemed to know what he is doing with math! I NEEDED you… where have you BEEN the last 3 years, tall dark and handsome? eh?

in addition to “playing” with the CORRECTLY STATED “re-solved” equation i made? to see what 2 lenses will MAKE when used as a compound lens?

How about… you start putting THIS idea into your (very powerful) ray tracing software, and seeing what you can get out of it? Up until NOW, you have been stuck with the notion of finding “just the right collimation lens” to stick behind “the cool single lens”.

I think but do not know… that when you talk about putting the “small asphere collimation lens” next to the LED emitter? i think what you are doing, is putting it at its focal length? Stop doing that… start LOOKING at the problem the proper way. Namely, that you are going to find two lenses focal length values, that combine to produce “one compound lens”.

the BFL ? will tell you for each play, how far away from the LED it is “set”. The “airgap” tells you the distance to the front lens…

I am not “smarter” than anyone else here, in fact? you are smarter and better educated than “me” at this, sunshine. You have a VERY POWERFUL TOOL, your ray tracing software? I just want you to start USING your (i admit it) superior math skills? and your superior tools at your disposal? To crack this.

Heres what will happen, if you just “pick” two lenses out of a hat? you WILL GET some advantage, but, at first blush, it wone be enough of an advantage, to overcome the slight loss of using a second lens. But, once you get “somewhere”, you can go back and “replace” one of those 2 lenses? with two other lenses.

your new THREE lens solution? will be the same focal length as the ORIGINAL single lens ! and each lens will shape a little of the beam down along the way… THAT overcomes the slight losses of using more lenses…

you ray tracing model will not be perfect in real life? but… “hand tuning” with a yardstick and “helping hands” holding lenses… will guide you with practical real world results to guide you quicker. cigarette smoke blown in to allow you to SEE whats going on, and moving a piece of white paper back and forth to measure the diameters of the cones of light? will tell you where you are “losing” at each small step thru the lenses.

YOU know “who” the brains on this site are? Who dont poop their pants when faced with a little algebra? you guys just need to start LOOKING at the problem from THIS point of view… its the keys to the kingdom, and its not as hard as it looks, once you start playing and hand tuning the solution…

for the record? i went to a small state university, that wasn’t very “highly rated” when i got out into my own professional field when young. Smug men were always showing off their “much better school” degrees and making fun of me. You know what? I programmed them “under a table” and fried them…

You guys are all standing around ina group, looking up a TREE, wondering how to get up to the upper branches? I come along, and i am looking at the tree from the other side? I have a COMPLETELY different VIEW of the “problem”…

“hey guys! we can climb up right over here! come look!”

you guys can build better than me, there are MANY guys way better than me at the math. Your just not looking at the problem fromthe correct point of view. THIS is the correct view.

Well, that didn’t really answer the question, but if you want you can download the ray tracer for yourself and try it out, it’s free and pretty cool.
Takes a while to figure out though.

Using 100mm as the focal length for two lenses and 133.34mm as the airgap the combined focal length would be 150.0150015mm

(100 * 100) / (100 +100 –133.34) = 10000/66.66

Using the same 100mm focal lengths and 133.34mm airgap the back focal length would be –50.0150015

(100 * (133.34 - 100)) / (133.34 - (100 + 100)) = 3334 / –66.66

boundless.com page on compound lenses

if you play with 2 lenses of equal FL ? (here, both 100mm FL)

start with 0mm airgap (touching) and do it a few times in steps of increasing gap? and stop just shy of 100mmm airgap.

a negative BFL? indicates you tried to do something impossible… here? with two 100mm FL lenses? you cant hit 100mm airgap nor go over it… with two 100mm FL lenses? stay below 100mm airgap….

each lens is doing SOME of the work, neither lens is doing ALL of the work.

you’ll see a RANGE of values from 50mm to close to 100mm as you do it in steps…

tekwyzrd? you are NOT making a “beam expander”, nor are you making anything even remotely resembling a “galilean telescope” setup… all you are doing? is combining 2 independent lenses, into one compound lens…

if the focal lengths are the same 100mm (easiest to start with) you have to keep the airgap under 100mm…

after a while you get bored trying different matching FL’s… i would suggest trying to play with one “littler one” and one “bigger one”… the littler one is always the collimating lens, and the bigger one is always the main objective. i think the convention is #1 is always the collimating and the #2 is always the main objective? but dont quote me…

and anytime you get a negative BFL ? you tried to do something impossible… this will not allow you to violate the laws of physics…

it IS going to mess with your head, at first? having two lenses, neither one at its own “focal length”, because each lens is doing PART of the job formerly done by the ONE lens…

i think this is whats irking Mr. Enderman… i think his “small asphere collimating lens” is set at its focal length in his ray trace of it, i believe anyways, i could be wrong.

think of it? as… with the traditional ONE LENS zoomie? you are getting further away from the paint can… this allows you to put half the paint can at the start and half the paint can a ways towards the final lens.

or? witha single lens zoomie? you are looking for this ONE HUGE GUY to carry a giant LOG down the trail? you run out of guys big enough and wide enough, they dont exist… so what do you do? you get 4 regular guys to line up and carry the log… each of 2 or more lenses is doing PART of the work, none of them is doing ALL the work.

hence, none of them are sitting at their focal length.

I know it wasn’t possible. Just trying to provoke a ‘Huh? What?’ reaction. :disappointed:

2x 100mm fl lenses = < 100mm
80mm and 100mm = < 80

if anmyone first start to “see it” and have the “eureka moment”?? my money is on tekwyzrd… i can clearly see his enthusiasm, and i can see his “sideways thinking” he is capable of…

how? he’s seeing what i saw, the weird way the equation sort of resembles the way we figure out parallel resistances and series capacitors in electronics. i dont know why it looks similar to that…

but, working with reciprocals is maddening… working with the resolved equation is much nicer…

I suppose one could re-solve the equation to isolate either the collimating lens or the main objective lens… i never had much use for it for my purposes…

but, now you know “how” this light was put together by the designer…

and its not HALF as mysterious to you as it was before…
you start with “2 lenses” replacing one lens? then the forward lens gets broken into 2 lenses… etc etc

when you use only 2 identical big fl lenses? you still get the “free lunch”, but… its just usually a twig of parsel;y on the side, unless you get a collimating lens working perfect. But when you get 3 or even 4 lenses working, each a little bit apart?? you start to get a “free double order of fires” with your burger, it becomes worth it, but, only once its “tuned” by hand in the build.

I’m honestly and truly? Trying to “give back” to the site the best and biggest gift i can bestow. Anyone else but me? would probably keep this a secret, and want to charge big money for THEIR builds they sold. I’m feeling guilty, that i am only “here” to learn to fabricate and then to get used to machining, and now foundry work? practicing on flashlights, so i can do my “real work” on another site?? I’m TRYING to give you the keys that allow you to start and drive a racecar instead of a regular commuter car or a big truck.

once you have played enough? you start to see the “practical rules of thumnb” using two lenses to create a new single compound lens… it becomes almost second nature…

It helps to have a calculator that can re-use calculations and touch where you want to change values. A good android calc is MathsApp